A numerical framework written in Julia that implements Langevin dynamics, used to thermodynamically sample a system of interacting particles consistent with the canonical ensemble. Implemented features include:
- Pairwise interactions: Lennard-Jones potential.
- Cell list: speeds up the force computation, making it O(N) instead of O(N^2).
- Stochastic integration: Euler-Maruyama.
- Periodic boundary conditions.
- Computation of kinetic energy, potential energy, instantaneous temperature.
- Histogram plot of velocity distribution.
Features to be implemented/known limitations include:
- Generalize it to noncubic domains.
- Implement other integrators/interaction models.
- Computation of more observables.
- Efficiency - the current version is far from optimal.
All simulation parameters are defined as constants in input.jl. To run the code, simply execute main.jl. A brief description of the other scrips is below.
cell_lists.jl: Initializes and unwraps neighbor lists to compute the force due to pairwise interactions.hamiltonian.jl: Computes kinetic/potential energy and the force.initialize.jl: Generates initial lattice and initial velocities.integrator.jl: Euler-Maruyama integrator.PBC.jl: Periodic boundary conditions.histogram.jl: Normalizes the Gibbs distribution and plots velocity histogram.
This code was written as a test code, as this was my first experience with Julia.
I am currently working on a larger project, which implements nonequilibrium Langevin dynamics for a series of background flows, for which this code serves as a basis/prototype.
All feedback is greatly welcome.